Search Extension

API/RSS

Social

From theory to practice: Can Based Rollup realize the L1 sorting-driven Rollup solution?

YBB Capital

24-06-20 20:00

Read this article in 23 Minutes

AI summary

View the summary

Original title: "From theory to practice: Can Based Rollup realize the L1 sorting-driven Rollup solution?"

Original author: Ac-Core, YBB Capital Researcher Ac-Core

Foreword:

The working principle of Ethereum is that each node stores and executes each transaction submitted by the user. In order to expand the entire network, Ethereum adopts the Rollup solution. Simply put, it moves most of the transaction processing to the off-chain (L2), thereby reducing the burden on the Ethereum main network (L1) and reducing transaction fees, that is, Rollup = a set of smart contracts on L1 + network nodes on L2, that is, on-chain smart contracts and off-chain aggregators. Its own settlement, consensus and data availability rely on Ethereum itself, and it is only responsible for executing transactions. L2 network nodes are composed of multiple parts, the most important of which is the sorter, but currently the sorters of Rollup are facing the problem of centralization.

Rollup and Sorter

Rollup is a scaling solution for Ethereum (L1) that executes transactions off-chain and packages them in blocks. For each block, Rollup publishes the data required to reconstruct the chain state (as a source of data availability) to the data availability layer, and publishes the proof of correctness of the off-chain execution to the settlement layer (there are two types of Rollup. In the case of ZK-rollup, zero-knowledge proofs are published for each block; in the case of Optimistic rollup, fraud proofs are only published in the event of a dispute). After EIP-4844, when data publishing switches to blobs, this layer may be called the "data publishing layer". The smart contract of the Rollup verifies the published proof on L1, and each Rollup will have one or more bridges to enable data transmission between chains as well as deposits and withdrawals.

In the implementation logic of Rollup, the Sequencer is a key component. It is mainly responsible for receiving transaction requests on L2, determining their execution order, and packaging transactions into batches (Batch) and finally transmitting them to the Rollup smart contract on L1, playing an important role in improving transaction processing efficiency and reducing costs.

The functions and working principles of the sequencer mainly include four parts.

1. Receiving transactions:The sequencer receives transaction requests from users or applications. These transactions are first processed on L2, rather than directly on the Ethereum mainnet;

2. Sorting transactions:The sorter is responsible for sorting the received transactions and determining the order in which they are executed. This process is similar to what Ethereum miners do before packaging transactions into blocks;

3. Packing transactions:The sorter packages the sorted transactions into batches, which contain summary information of multiple transactions;

4. Submit to L1:Finally, the sorter submits the packaged transaction batches to the Ethereum mainnet (L1) for settlement and data storage on the mainnet. In this way, L1 can verify and store state updates on L2.

Although Rollup technology provides an effective expansion solution, there are still some problems in the design and implementation of the sorter. The first one is the centralization problem. Most Rollup projects currently rely on centralized sorters, which are usually controlled by a single entity or a few entities. The intuitive risks brought are lack of transparency and single point of failure.

Apart from the above rigid explanation, the discussion on the L2 decentralized sorter solution, whether it is the L2 solution of sorting pool node pledge directly released by Metis to the market, or the independent project Espresso solution mentioned in previous articles of YBB Capital, is essentially the distribution of the sorting "profit cake" and the expectation of future market speculation space, so the correctness of interests and orthodoxy is the key that is difficult to avoid.

Historical background and design of Based Rollup

Image source: @drakefjustin

The concept of Rollup was first proposed by Vitalik Buterin, the founder of Ethereum. Its original idea was to achieve a completely unconstrained "Total Anarchy" state to allow anyone to expand transactions without restrictions. In view of the above problems with the current sorter, in 2023 Ethereum Researcher: Justin Drake proposed a solution based on Rollups to manage the sorter by Ethereum L1 itself, the content of which is as follows (source see extended link 1):

Definition:

"When the sorting of the rollup is driven by the base layer (L1), we call it L1-based or L1-sorted rollup. Specifically, L1-based rollup means that the next L1 proposer can cooperate with L1 searchers and builders to include the next Rollup block in the next L1 block without permission."

Advantages:

● Liveness: Based Rollup enjoys the same liveness guarantee as L1. Note that non-Based Rollups with escape hatches will have reduced activity (escape hatches are a safety mechanism in Rollups that allow users to safely withdraw assets from L2 back to the L1 main chain when problems occur in the Rollup system. It is similar to an emergency exit);

Weaker settlement guarantees: transactions in escape hatches must wait for a timeout before settlement is guaranteed;

Censorship-based MEV: Rollups with escape hatches are vulnerable to adverse MEV effects from short-term sorter censorship during the timeout;

Network effects are at risk: large-scale exits triggered by sorter activity failures (such as a 51% attack on a decentralized PoS sorting mechanism) will destroy the network effects of Rollups. Note that unlike L1, Rollups cannot use social consensus to recover gracefully from sorter activity failures. Mass exit is the sword of Damocles in all known non-Based Rollup designs; Gas penalty: Transactions settled through escape pods often incur gas penalties for their users (e.g. due to suboptimal data compression when transactions are not batched). ● Decentralization: Based Rollup inherits the decentralization of L1 and naturally reuses the infrastructure of L1 searchers-builders-proposers. L1 searchers and builders are incentivized to include rollup blocks in their L1 blocks to extract the MEV of the rollup. This in turn incentivizes L1 block proposers to package rollup blocks on L1. ● Simplicity: Based Rollup sorting is the simplest, even much simpler than centralized sorting. Based Rollup does not require verification of the sorter signature, does not require an escape hatch, and does not require external PoS consensus.

Historical note: In January 2021, Vitalik called the L1-based sorting scheme "complete anarchy", which risks submitting multiple rollup blocks at the same time, resulting in a waste of Gas and work. The current block proposer-builder separation scheme (Proposer-Builder Separation, PBS) can strictly control the L1 sorting, with at most one rollup block per L1 block and no gas waste. When the n+1 block of the rollup (or n+k for k >= 1) contains a SNARK proof of block n, the waste of ZK-rollup proof work can be avoided.

● Cost: Based Rollup has zero gas costs—it doesn’t even need to verify signatures from decentralized or centralized sorters. The simplicity of Based Rollup reduces development costs, shortens release time, and reduces the exposure of code vulnerabilities. Based Rollup’s sorting is also token-free, avoiding the regulatory burden of token-based sorters.

● Alignment with L1 economy: MEVs originating from Based Rollup naturally flow to the L1 on which it is based. This flow strengthens L1 economic security and, in the event of MEV destruction, increases the economic scarcity of L1 native tokens. This close economic alignment with L1 may help build the legitimacy of Based Rollup. Importantly, despite sacrificing MEV revenue, Based Rollup retains the option to generate revenue from L2 congestion fees (such as L2 base fees in the form of EIP-1559).

● Sovereignty: Despite delegating sorting to L1, Based Rollup retains sovereignty. Based Rollup can have a governance token, charge basic fees, and can use the proceeds of these basic fees when appropriate (such as Optimism to fund public goods).

Disadvantages:

● No MEV income: Based Rollup gives up MEV to L1, limiting its income to basic fees. Counterintuitively, this may increase the total income of Based Rollup. The reason is that the rollup landscape seems to be winner-takes-all, and the winning rollup may use the security, decentralization, simplicity, and consistency of Based Rollup to achieve dominance and ultimately maximize revenue.

● Constrained sorting: Delegating sorting to L1 reduces sorting flexibility. This makes certain ordering services more difficult, or possibly even impossible:

○ Preconfirmation: Fast preconfirmation is not a problem for centralized ordering, and can be achieved with external PoS consensus. Fast preconfirmation with L1 ordering is an open problem with many promising research directions, including EigenL, Inclusion Lists, and Builder Bonds.

○ First Come, First Served (FCFS): It is uncertain whether Arbitrum-style FCFS ordering can be implemented on Based Rollup. EigenL may provide an FCFS overlay for L1-ordered Based Rollup.

Naming:

The name "Based Rollup" comes from its proximity to the base chain (Base L1). This conflicts with Coinbase's recently announced Base chain, which is a strange coincidence. In fact, Coinbase shared two design goals in their Base announcement:

● tokenlessness: “We have no plans to issue a new network token.”

● decentralisation: “We [...] plan to gradually decentralize the blockchain over time.”

Base can achieve tokenless decentralization by becoming Based Rollup.

Image source: @jchaskin22

In theory, Based Rollup allows anyone to extend to Rollup blocks, publish the sorted transaction state changes to L1, and extract MEV from L2, so that all sorting and security are provided by Ethereum L1. This can circumvent the external proof-of-stake consensus and the specific Rollup token requirements. At the same time, compared to other Rollups that are essential for keeping assets safe, the "emergency escape hatch" function can be removed in the vision of Based Rollup. The process only needs to ensure the safe operation of Ethereum, and transactions on Rollup can be completed smoothly.

Other core structures:

Competitive Rollup (BCR - Base Competitive Rollup)

is an innovative blockchain expansion solution developed by Taiko Labs. BCR aims to improve the efficiency and security of Rollup through a competitive mechanism, allowing different participants to freely compete to submit blocks and generate proofs, thereby improving the performance and decentralization of the entire network. The summary is as follows.

● Features

Open competition: Allows any qualified participant to compete to submit blocks and generate proofs. This open mechanism reduces centralized control and improves the decentralization of the network. Competitors win rewards and transaction fees by providing better services;

Efficient expansion: It can effectively improve the efficiency of block generation and verification. Multiple competitors can work in parallel, avoiding single-point bottlenecks, improving the speed of transaction processing and the expansion capacity of the network;

Security: Through the competition mechanism involving multiple parties, the system's anti-attack ability is improved. Blocks and proofs generated by multi-party competition increase the transparency and security of the system, making it difficult to be controlled or attacked by a single entity.

● Advantages

EVM compatibility:BCR is fully compatible with the Ethereum Virtual Machine (EVM), allowing existing Ethereum smart contracts and decentralized applications (DApps) to be easily migrated to run on BCR without major modifications;

High throughput:Since BCR can process transactions and generate blocks in parallel, the network's throughput is significantly improved, capable of handling higher transaction volumes, and reducing transaction costs and delays;

Decentralization:Through decentralized block generation and proof mechanisms, the decentralized nature of the network is ensured, reducing the risk of being controlled by centralized entities.

● Disadvantages

Increased complexity:The complexity of the system requires complex algorithms and protocols to coordinate block generation and verification among multiple competitors. The complexity of smart contracts may require additional logic to handle competition results;

Potential problems:Under the competition mechanism of BCR, when multiple competitors calculate and submit blocks at the same time, the cost may rise. Users may face higher transaction fees when using BCR for transactions, especially when the network is busy or the competition is fierce. At the same time, large nodes with abundant resources may have an advantage in the competition, resulting in centralized concentration.

Booster-based Rollup

Booster is a special participant in BBR. It is responsible for optimizing transaction batches, compressing transaction data and processing multiple transaction batches in parallel. Its actual role is to separate execution and storage, keep L2 execution while L1 decentralized, and keep the address of smart contracts on L1 and all BBRs consistent.

At the same time, it also faces challenges such as increased system complexity, resource consumption, and potential centralization. In the future, BBR still needs to be further optimized and expanded to meet the needs of the evolving blockchain technology.

Image source: Taiko Labs

Conclusion:

From the overall observation, Based Rollup can be regarded as a huge change in the way Ethereum's second-layer expansion is carried out. The sorting of Rollup is directly entrusted to the first-layer proposer, and the separation design of the proposer and the builder is used to let the first layer perform all the sorting roles for Rollup. At the same time, MEV is also expanded, so that L2 searchers can send transaction bundles to L2 builders, who are also L1 searchers, and these complete L2 blocks then become part of the L1 blocks, and are ultimately processed by L1 builders and the Ethereum mainnet.

Whether Based Rollup can be assumed to be the final solution to Rollup remains to be verified, but it is undeniable that it is a major innovation in Ethereum's second-layer expansion at this stage, providing a safer and decentralized solution for expansion. If we bring the same thinking into the Bitcoin ecosystem, the difficulty of implementation is understandable compared to the native and decentralized VM implemented by Ethereum with strong scalability, so the entire industry still has a long way to go in solving the problem of truly decentralized expansion.

Extended link:

【1】 Based
rollups—superpowers from L1 sequencing https://ethresear.ch/t/based-rollups-superpowers-from-l1-sequencing/15016/1

【2】The different types of ZK-EVMs https://vitalik.eth.limo/general/2022/08/04/zkevm.html

Reference article:

【1】The game of credit: Rollups controlled by multi-signatures and committees https://mp.weixin.qq.com/s?__biz=Mzk0OTYwMDM1Mg==&mid=2247486544&idx=1&sn=7a0ee8eb1d32e6ef0bd08354986c0a1f&source=41#wechat_redirect